The long-term goal is to provide for an improved immunoassay as an adjunct for the diagnosis of Lyme disease. The assay preferably would have (a) greater sensitivity for detecting infection during early disease, (b) have as a good if no better specificity than currently available assays, alone or in combination, for all stages of disease, and (c) be sufficiently informative that assay interpreters infer the strain of B. burgdorferi a patient is infected with. Specific aims for this R21/R33 application are the following: (1) Use a genome-wide proteome array to interrogate IgM responses of patients with LD and experimental animals infected with B. burgdorferi. The approach will be similar to what was successfully used for studies with an array of IgG responses of humans and mice to this pathogen. The hypothesis that sensitivity of IgM assays is heightened by including two or more OspC type will be tested. (R21 phase) (2) Investigate whether profiles of antibody reactivity to sub-unit antigens provide added-value for inference about the strain of B. burgdorferi someone has been infected with. This will be evaluated in experimental animals infected with different strains and with sera from patients for whom the infecting strain was recovered. Hypotheses that (a) OspC type-specific responses can be distinguished and (b) patterns of reactivity to the BBK07 and BBK12 proteins vary with the infecting strain will be tested. (R21 phase) (3) Develop an immunoassay for eventual clinical testing by (a) selecting the most informative set of recombinant antigens for achieving a high predictive value while minimizing cost, and (b) selecting the most suitable platform, provisionally from among ELISA, immunostrips, and Luminex, for implementation in clinical research evaluation of the assay. The choice of antigens will be from those identified in a previous study, as well as from the results of aims 1 and 2. (R33 phase)